gents, thanks for the fascinating discussion. All things I need to know.
This amp design is meant to be for common folk, so I want easy to find tubes. The original kit this design came from used 6p3p output tubes and a 6N1 driver and the 'T7 seemed to be closest, but I'd love to do a version with the older 6SN7 tube. The idea is to build an amp that is simple, easy to build, and silicone-free. 12AT7 is common as can be, and 6l6GC is too. I messed with some Soviet tubes on the lower-power Kit, with a b+ of 320v, but the transformer on this one will red-plate those 6p3p-type tubes that are closer to OG 6L6 in spec, can't take 350v to the plates.
Anyway, I still haven't reached 200v to the plates on this job, ran out of time. A 4.7k ohm resistor led to 343v at supply, 184v at the plates and 1.9v at the cathodes. I've got a free tmrw so I'll be plugging away.
thanks!
This amp design is meant to be for common folk, so I want easy to find tubes. The original kit this design came from used 6p3p output tubes and a 6N1 driver and the 'T7 seemed to be closest, but I'd love to do a version with the older 6SN7 tube. The idea is to build an amp that is simple, easy to build, and silicone-free. 12AT7 is common as can be, and 6l6GC is too. I messed with some Soviet tubes on the lower-power Kit, with a b+ of 320v, but the transformer on this one will red-plate those 6p3p-type tubes that are closer to OG 6L6 in spec, can't take 350v to the plates.
Anyway, I still haven't reached 200v to the plates on this job, ran out of time. A 4.7k ohm resistor led to 343v at supply, 184v at the plates and 1.9v at the cathodes. I've got a free tmrw so I'll be plugging away.
thanks!
I was going to suggest trying that. I would have expected a higher voltage at the 12AT7 cathode, but that's probably good. Are the 470R and 100R from cathode to ground unchanged? If so, that means only 3.3mA of plate current through the 12AT7. That's fine. If you hear more hum, try replacing the 22uF decoupling capacitor with 47uF 450V.
If you have 2.7k or even 2.2k, and you can increase the value of the decoupling capacitor to 47uF or 68uF, then you can get better drive from the 12AT7. The ideal spot would be about 190V to 200V on the plate, 2V at the cathode, about 3.5mA plate current.
12AT7 and 6L6GC are often stocked for guitar amps at musical instrument stores, so they'll be easily available for a long time to come. I can see choosing those for that easy availability.
Last edited:
Hey thanks for all that. Good stuff. I do have those resistors, and more 47uF caps, don't think I have any 68uF. Would it be overkill to throw another R-C stage in, say 22Kohm-22uF, before the Driver Supply stage, which will be 2.2Kohm-47uF ? Guess I'll try this and see if there's any hum before I go that way.
You have to calculate how much current will go through any decoupling resistor (in series with the voltage), which will tell you how many volts will be burned off by that resistor.
If I understand you correctly, you're asking if you should add another RC decoupling stage in the 12AT7 plate supply? If so, no, because Ohm's Law tells us that current through a resistor will decrease both the current and voltage to the plate supply of the 12AT7.
The plate current (Ip) the 12AT7 draws will pull current through the additional 22k ohm resistor, so voltage will be dropped across that resistor.
The reduction of voltage to the 12AT7 plate will reduce Ip in the 12AT7.
You have 570 ohms in the 12AT7 cathode, for the cathode bias. With let's say 2mA Ip, you'll be back to only 1.14V at the cathode, and the 12AT7 may be back in grid current territory.
You want more plate voltage to the 12AT7, not less.
100uF 450V capacitors are not expensive and are easy to find, or you can connect two 47uF capacitors in parallel to get 94uF (close enough). If you put a 2.2k (2200 ohms) resistor instead of 4.7k, you could replace the 22uF with larger caps like that if you need to.
--
If I understand you correctly, you're asking if you should add another RC decoupling stage in the 12AT7 plate supply? If so, no, because Ohm's Law tells us that current through a resistor will decrease both the current and voltage to the plate supply of the 12AT7.
The plate current (Ip) the 12AT7 draws will pull current through the additional 22k ohm resistor, so voltage will be dropped across that resistor.
The reduction of voltage to the 12AT7 plate will reduce Ip in the 12AT7.
You have 570 ohms in the 12AT7 cathode, for the cathode bias. With let's say 2mA Ip, you'll be back to only 1.14V at the cathode, and the 12AT7 may be back in grid current territory.
You want more plate voltage to the 12AT7, not less.
100uF 450V capacitors are not expensive and are easy to find, or you can connect two 47uF capacitors in parallel to get 94uF (close enough). If you put a 2.2k (2200 ohms) resistor instead of 4.7k, you could replace the 22uF with larger caps like that if you need to.
--
Understood, thanks.
I've just installed a 47u cap and 2.7Kohm, let's throw the switch....looks like I'm just kissing 189v at the plates, 354v at the supply cap, 2.05v at the bias. Gonna try that 2.2Kohm.
I've just installed a 47u cap and 2.7Kohm, let's throw the switch....looks like I'm just kissing 189v at the plates, 354v at the supply cap, 2.05v at the bias. Gonna try that 2.2Kohm.
B+ = 354V
Vp = 189V
Vk = 2.05V
Rp = 47k
So...
354V - 189V = 165V
165V/47000R = 0.00351A (3.5mA)
2.05V/570R = 0.0036A (3.6mA -- probably due to another 100uA from the NFB)
Looks OK to me. You should be able to swing 50V peak reasonably cleanly. That gives you some headroom for the NFB, but not gobs of it.
Vp = 189V
Vk = 2.05V
Rp = 47k
So...
354V - 189V = 165V
165V/47000R = 0.00351A (3.5mA)
2.05V/570R = 0.0036A (3.6mA -- probably due to another 100uA from the NFB)
Looks OK to me. You should be able to swing 50V peak reasonably cleanly. That gives you some headroom for the NFB, but not gobs of it.
Attachments
Last edited:
ok gang, I'm sitting here jamming out to Peg and I'm very happy. Right now there's a 1Kohm resistor and a 47uF cap as driver supply at 367v, with 195v at the plates and 2.14v at the cathodes. Here's an updated schematic:
it sounds great, but I'm still getting bad distortion from the full-strength signal coming from my phono pre. It's a DIY 834p with a pretty hot output, it caused clipping with the SS amp I was using when I started all this, and it has a 50K pot on the output. I'm getting great sound from this system with that pot turned down about 50%, so I still have a very low input sensitivity that will be a problem with lots of gear. Should I put 1k resistors on the input pins for the 12at7 or ...?
it sounds great, but I'm still getting bad distortion from the full-strength signal coming from my phono pre. It's a DIY 834p with a pretty hot output, it caused clipping with the SS amp I was using when I started all this, and it has a 50K pot on the output. I'm getting great sound from this system with that pot turned down about 50%, so I still have a very low input sensitivity that will be a problem with lots of gear. Should I put 1k resistors on the input pins for the 12at7 or ...?
Are you sure your 834P preamp is not clipping? If that clips, the power amp can only amplify that clipped input signal.
I'm not sure how a couple of 12AX7s can make a hot enough output to overload the 12AT7's input (grid). Do you have a step-up transformer at the input of the 834P?
Does the amp distort when fed from the output of a CD player or DAC?
If the problem is that the amp is too sensitive and clips with much less than 1V input signal, you can make the amp less sensitive, lower its distortion, and decrease its output impedance by reducing the value of the 2k ohm negative feedback resistor (from the output transformer's 8 ohm tap to the top of the 100 ohm resistor in the 12AT7 cathode circuit). Can you clip a resistor in parallel to that 2k ohm resistor? Perhaps start with 6.8k ohms in parallel (total 1.55k), see how that works. You could even go as low as 3.3k in parallel with 2k (total 1.2k).
There are other things you could try. Perhaps increase the value of the 470 ohm resistor in the 12AT7 cathode to 560 ohms, or even 680 ohms. Then see what the plate current is through the 12AT7. 3mA is a good target, although a little more (like you have now) is fine too.
Of course, it might be that the input stage has too much gain and you should use a twin-triode there that has less gain, like the 6FQ7 Chris suggested. Another one you could try is 12AU7 (although some people will get all upset because "it distorts too much"). At least 12AU7 is pin-compatible with 12AT7, so it would be a lot less work for you to try it.
Ah, another pin-compatible twin-triode you can try is the 12AV7 or its better version, 5965. Jut pop one of those in there, take measurements, listen, see what you get. Those have less gain (lower mu) than 12AT7, but more gain than 12AU7.
Yet another thing you could try is take the 220R 1W resistor you have going from the 6L6 pin 4 to the UL tap on the output transformer (OPT) primary and connect it to the 6L6 plate (pin 3) instead (i.e., wire the 6L6 to triode mode). That will decrease gain, and thus reduce the overall sensitivity of the amp. It will also reduce the power output by about half, which you may not want.
Just some untested ideas that maybe can help with the sensitivity problem. Maybe.
--
I'm not sure how a couple of 12AX7s can make a hot enough output to overload the 12AT7's input (grid). Do you have a step-up transformer at the input of the 834P?
Does the amp distort when fed from the output of a CD player or DAC?
If the problem is that the amp is too sensitive and clips with much less than 1V input signal, you can make the amp less sensitive, lower its distortion, and decrease its output impedance by reducing the value of the 2k ohm negative feedback resistor (from the output transformer's 8 ohm tap to the top of the 100 ohm resistor in the 12AT7 cathode circuit). Can you clip a resistor in parallel to that 2k ohm resistor? Perhaps start with 6.8k ohms in parallel (total 1.55k), see how that works. You could even go as low as 3.3k in parallel with 2k (total 1.2k).
There are other things you could try. Perhaps increase the value of the 470 ohm resistor in the 12AT7 cathode to 560 ohms, or even 680 ohms. Then see what the plate current is through the 12AT7. 3mA is a good target, although a little more (like you have now) is fine too.
Of course, it might be that the input stage has too much gain and you should use a twin-triode there that has less gain, like the 6FQ7 Chris suggested. Another one you could try is 12AU7 (although some people will get all upset because "it distorts too much"). At least 12AU7 is pin-compatible with 12AT7, so it would be a lot less work for you to try it.
Ah, another pin-compatible twin-triode you can try is the 12AV7 or its better version, 5965. Jut pop one of those in there, take measurements, listen, see what you get. Those have less gain (lower mu) than 12AT7, but more gain than 12AU7.
Yet another thing you could try is take the 220R 1W resistor you have going from the 6L6 pin 4 to the UL tap on the output transformer (OPT) primary and connect it to the 6L6 plate (pin 3) instead (i.e., wire the 6L6 to triode mode). That will decrease gain, and thus reduce the overall sensitivity of the amp. It will also reduce the power output by about half, which you may not want.
Just some untested ideas that maybe can help with the sensitivity problem. Maybe.
--
Last edited:
I made a mistake in the verbal description of how to wire the 6L6 into triode. I edited my last post to fix it. You might want to read that part again so you don't get lead astray by my dumb mistake. 🙂
I actually have the amp switchable between U/L and Triode Strapped. Triode is really nice. But wiring the screens to b+2 is just too much, man.
Sorry, that was a mistake. I didn't mean wire the screen to B+. That would actually make the amp more sensitive, by wiring the 6L6 into pentode (but incorrectly as I had described it).
Does the 834P output crunch the amp even if you have it switched to Triode?
Does the 834P output crunch the amp even if you have it switched to Triode?
yup, same crunch. I'm back from voting, just beat the rain, gonna try some bigger cathode resistors, see what happens.
Does the 834P output crunch into any amp you connect it to? Or does it only crunch when connected to this 6L6 amp?
When I finally got my 834 working, I realized that some records were clipping with my Kenwood AV receiver. Loud trumpet and piano passages, stuff like that, would just get mashed up and lose all detail. Adding a 50k pot between the PCB and the RCA jacks let me turn it down just enough to remove that distortion, at a cost of some of the bottom end.
Now, I use a 6p3p power amp, a DIY kit that I'm copying for this project. One variant I've built uses the 6N1 as driver as spec'd, another one has a 12AT7 in that spot with the only change being that cathode resistor, it's 1k with the 6n1 and 470R with the 12AT7. That amp uses a smaller transformer with a b+ of about 320v Neither of those amps have any distortion with this preamp, always have it wide open, so I kinda figured they had a higher input ceiling than the AV receiver I had been using. Or the tube is set up correctly for a line-level signal.
anyway, I put 680ohm resistors on the Cathodes and fired it up. b+ had raised a bit, to 383v, and c1 was 414v, just a bit too high: the choke has a max rating of 400v and I'm right there. Changing the driver supply resistor back to 2.7kohm got my c1 to 408v, Choke is seeing 398v, b+ is 378v, driver supply is 361v, driver plates see 210v and cathodes see 2.54v.
and the distortion. still. no real change, I have to turn the 834 down at least 1/3rd, maybe half, which is way more than I was needed with that SS amp. I'm stumped.
updated the sch:
Now, I use a 6p3p power amp, a DIY kit that I'm copying for this project. One variant I've built uses the 6N1 as driver as spec'd, another one has a 12AT7 in that spot with the only change being that cathode resistor, it's 1k with the 6n1 and 470R with the 12AT7. That amp uses a smaller transformer with a b+ of about 320v Neither of those amps have any distortion with this preamp, always have it wide open, so I kinda figured they had a higher input ceiling than the AV receiver I had been using. Or the tube is set up correctly for a line-level signal.
anyway, I put 680ohm resistors on the Cathodes and fired it up. b+ had raised a bit, to 383v, and c1 was 414v, just a bit too high: the choke has a max rating of 400v and I'm right there. Changing the driver supply resistor back to 2.7kohm got my c1 to 408v, Choke is seeing 398v, b+ is 378v, driver supply is 361v, driver plates see 210v and cathodes see 2.54v.
and the distortion. still. no real change, I have to turn the 834 down at least 1/3rd, maybe half, which is way more than I was needed with that SS amp. I'm stumped.
updated the sch:
I don't know, it sounds like there might be something wrong with the phono preamp. Perhaps DC on its output? DC offset?
Do digital sources crunch the 6L6 amp input too?
What cartridge are you using into the 834P?
Does your 834P have an input transformer?
If you have a signal generator (will need a voltage divider for use with the 834P) and an oscilloscope, or an RMS voltmeter, you could send a 1mV 1kHz sine wave into the 834P input and measure the signal voltage at the output of the 834P. That would tell you how much gain it has.
If the gain is so high that you see 1.5V or more out with 5mV in, then it could be that your 834P is set to too high gain, and is overloading everything downstream, especially if you're using a higher output cartridge like a Shure M44xx, M35X, SC35C, etc.
Strange for a phono pre to blast the downstream inputs like that. The usual problem is that the phono preamp puts out less signal volts than the DAC, CD player, etc. The problem I have is that the signal from my vinyl setup is lower than the signal from my DAC, so I have to adjust the volume controls every time I switch sources. I had to add a little gain stage (2X gain) to get the phono to be at about the same level as the digital. You seem to have completely the opposite problem going on.
--
Do digital sources crunch the 6L6 amp input too?
What cartridge are you using into the 834P?
Does your 834P have an input transformer?
If you have a signal generator (will need a voltage divider for use with the 834P) and an oscilloscope, or an RMS voltmeter, you could send a 1mV 1kHz sine wave into the 834P input and measure the signal voltage at the output of the 834P. That would tell you how much gain it has.
If the gain is so high that you see 1.5V or more out with 5mV in, then it could be that your 834P is set to too high gain, and is overloading everything downstream, especially if you're using a higher output cartridge like a Shure M44xx, M35X, SC35C, etc.
Strange for a phono pre to blast the downstream inputs like that. The usual problem is that the phono preamp puts out less signal volts than the DAC, CD player, etc. The problem I have is that the signal from my vinyl setup is lower than the signal from my DAC, so I have to adjust the volume controls every time I switch sources. I had to add a little gain stage (2X gain) to get the phono to be at about the same level as the digital. You seem to have completely the opposite problem going on.
--
Your amplifier has input level controls. It's not possible to overload the amplifier without clipping it.
If you're insisting that the amplifier not clip, without turning down the input volume controls, no matter how large the input signal, stop that. That's what volume controls are for.
If, instead, the issue is the nature of the amplifier's clipping behavior, it could be improved considerably by reducing the overlarge coupling cap. Something in the 0u047 to 0u1 F range will let the amplifier recover from clipping much faster (although it will still clip at the same level).
All good fortune,
Chris
If you're insisting that the amplifier not clip, without turning down the input volume controls, no matter how large the input signal, stop that. That's what volume controls are for.
If, instead, the issue is the nature of the amplifier's clipping behavior, it could be improved considerably by reducing the overlarge coupling cap. Something in the 0u047 to 0u1 F range will let the amplifier recover from clipping much faster (although it will still clip at the same level).
All good fortune,
Chris
I think the distortion was a result of the 834 phono preamp having a very high output. Someone over on reddit used the term Flat Topping. It sounds fine with a bluetooth source and I'll try it with another phono preamp a little later today. My 834 has a 12ax7 in the v3 position, many recommend a 'u7 for v3.
Now I'm wondering why it is that the kit amp I put together didn't have a problem with the hot signal from the 834. It's b+ is about 80v lower, that might have something to do with it.
Now I'm wondering why it is that the kit amp I put together didn't have a problem with the hot signal from the 834. It's b+ is about 80v lower, that might have something to do with it.
I still don't hear you talking about using your amplifier's input volume controls. Have read this in other threads, and it must be a very modern idea. Today's heavily processed and level controlled sources deliver a consistent output voltage, and everything gets played as loud as possible anyway, so why have a volume control?
All good fortune,
Chris
All good fortune,
Chris
well there's only the one volume control on the front of the amp, a 100k pot, and it stays as low as I can manage so I don't get in trouble with my neighbors.
I've been vague with my setup. I'm using a turntable , a DIY phono preamp and a DIY tube power amp. All made by me in the last year. I've built a couple different phono preamps around DIY EAR834p PCBs, there's a great thread over at LencoHeaven https://www.lencoheaven.net/forum/index.php?topic=26658.0 that covers the build. I use a 50k pot on the output because I noticed that certain records, mostly old jazz records, were too "hot" and I'd hear some kinda distortion, clipping, I'm not sure what term to use, but having that 50k in there and dialing it back 10% made that nasty sound go away. I don't have a 'scope so I can't measure the output.
When I got my first DIY tube power amp working, You are being redirected... I noticed that it could take the signal from my 834 pre at full power. So I just assumed it was cool. assumptions, amirite? and it actually sounded better, more bottom end, so I figured that was the way to go.
Now that amp has a 100k pot on the front end for volume control, here's the sch
and I started tinkering with that design. I wanted to rebuild it with off-the-shelf parts available here, so I switched the 6L6GC for the 6p3p and the 12AT7 for the 6N1. Went with the 'T7 because it has less gain than the 'X7. And because TubeLab uses it, TBH. The only change was the cathode resistor from 1k to 470R
This is an improved version of the design I found here on DiyAudio. Sounds great, easier on the tube with the added g2 filter stage, and again, had no problems with the signal coming from my 834. So I've built 2 versions of this kit, one with the stock 6n1 driver, one with the 12at7, both work great. So I built one from scratch:
This amp uses a Hammond 290AX that puts out a little more voltage, 650v, and I see 408v at C1. I tweaked the power supply a bit and added a resistor before the Choke (because the 158Q is only rated to 400v) and b+ is 378v. I've done away with the g2 filter because pentode mode is too frickin loud, the OPTs have U/L taps and it's switchable btw that and triode strapped.
So here's where it stands now. it sounds good with a bluetooth analog source, and it sounds good with my 834, as long as it's turned way down. I hooked it up to a Shure m65 clone I built, and it was thin and anemic, no bass and too bright. That m65 sounds good on my other tube amp, but sounds wimpy on my Kenwood. another mystery.
I've been vague with my setup. I'm using a turntable , a DIY phono preamp and a DIY tube power amp. All made by me in the last year. I've built a couple different phono preamps around DIY EAR834p PCBs, there's a great thread over at LencoHeaven https://www.lencoheaven.net/forum/index.php?topic=26658.0 that covers the build. I use a 50k pot on the output because I noticed that certain records, mostly old jazz records, were too "hot" and I'd hear some kinda distortion, clipping, I'm not sure what term to use, but having that 50k in there and dialing it back 10% made that nasty sound go away. I don't have a 'scope so I can't measure the output.
When I got my first DIY tube power amp working, You are being redirected... I noticed that it could take the signal from my 834 pre at full power. So I just assumed it was cool. assumptions, amirite? and it actually sounded better, more bottom end, so I figured that was the way to go.
Now that amp has a 100k pot on the front end for volume control, here's the sch
and I started tinkering with that design. I wanted to rebuild it with off-the-shelf parts available here, so I switched the 6L6GC for the 6p3p and the 12AT7 for the 6N1. Went with the 'T7 because it has less gain than the 'X7. And because TubeLab uses it, TBH. The only change was the cathode resistor from 1k to 470R
This is an improved version of the design I found here on DiyAudio. Sounds great, easier on the tube with the added g2 filter stage, and again, had no problems with the signal coming from my 834. So I've built 2 versions of this kit, one with the stock 6n1 driver, one with the 12at7, both work great. So I built one from scratch:
This amp uses a Hammond 290AX that puts out a little more voltage, 650v, and I see 408v at C1. I tweaked the power supply a bit and added a resistor before the Choke (because the 158Q is only rated to 400v) and b+ is 378v. I've done away with the g2 filter because pentode mode is too frickin loud, the OPTs have U/L taps and it's switchable btw that and triode strapped.
So here's where it stands now. it sounds good with a bluetooth analog source, and it sounds good with my 834, as long as it's turned way down. I hooked it up to a Shure m65 clone I built, and it was thin and anemic, no bass and too bright. That m65 sounds good on my other tube amp, but sounds wimpy on my Kenwood. another mystery.